To the Editor:
Multiple myeloma (MM) is an incurable hematologic malignancy with disparities in outcomes noted among racial-ethnic subgroups [1, 2]. A number of prior studies comparing overall survival (OS) of European Americans (EA) and African Americans (AA) with MM have found disparities, with lower OS among AA tied, in part, to worse socioeconomic factors among AA leading to less access to the novel treatments [3–5]. In a surprising rebuttal of previous findings, a recent study from Veterans Affairs hospitals showed that AA with MM have a superior OS than EA when all patients have equal access to novel therapies and transplant [6]. Their report is unique given the potential elimination of racial differences in drug affordability and healthcare access among the study population. The investigators examined patients in two groups: those 65 years and older and those younger than age 65. Even when doing this, AA had better OS; they showed that AA patients diagnosed before turning 65 had a significantly better median OS (7.07 years; 95% CI, 6.36–7.70 years) compared with EA (5.83 years; 95% CI, 5.44–6.09 years). At age 65 and beyond, median OS was similar between AA and EA. In another study, a meta-analysis of upfront MM clinical trials over more than two decades showed similar OS between enrolled AA and EA [7]. These two studies together suggest that unequal access to effective anti-myeloma therapy may partially explain the previously reported discrepancies between EA and AA with MM.
The contribution of early mortality (EM) on OS has not been assessed in any of these studies. EM in MM is usually attributed to advanced disease causing organ failure and comorbidities, while late mortality is mostly caused by emerging biologically resistant disease. EM maybe underreported in frontline MM clinical trials, given the selection bias towards enrolling younger patients with good performance status, adequate organ function, and favorable socioeconomic factors. In addition, race-based EM assessment is difficult in MM clinical trials because only ~5% of enrolled patients are AA, while they represent 20% of all MM patients in the US [8]. Therefore, it maybe instructive to identify the influence of EM on the outcome of AA and EA in the real-world setting.
In the present study, we have used the national cancer database (NCDB), which is the largest cancer database in the US including more than 70% of all newly diagnosed cancer patients, to investigate the contribution of EM on the long-term survival of MM patients stratified by race and age. We reviewed patients who were diagnosed with MM between 2004 and 2014 and reported to the NCDB, using the International Classification of Diseases for Oncology (ICD-O) code of 9732. AA and EA who underwent induction with chemotherapy and/or immunomodulatory imide drugs (IMiD) within 90 days of diagnosis and had a follow-up of ≥6 months after the diagnosis were included for analyses. We defined EM as death ≤6 months from diagnosis [9]. OS was calculated from the date of MM diagnosis, estimated by the Kaplan–Meier method, and compared by the log-rank test. All reported p values are two-sided. Statistical calculations were made using R statistical software version 3.4.0.
We identified 76,878 MM patients, of which 61,171 (80%) were EA and 15,707 (20%) were AA (Table 1). Overall, EM occurred in 8783 patients (11%), with rate of EM 10.3% (95% CI: 9.8– 10.7) for AA and 11.7% (95% CI: 11.5–11.9) for EA (p < 0.0001). The racial gap in EM widened in elderly patients aged ≥65 years with rates of 13.8% (95% CI: 13.0–14.7) in AA and 15.7% (95% CI: 15.3–16.1) in EA (p < 0.0001). In contrast, EM rate among younger patients aged <65 years at diagnosis was higher in AA (7.4%, 95% CI: 6.8–7.9) compared with EA (6.5%, 95% CI: 6.2–6.8) (p < 0.0001). EM rate has decreased over the years from 2004–2009 to 2010–2014 in cohorts of elderly as well as young patients (Fig. 1a, p < 0.0001 for all). EA more often received IMiD (19% vs. 18%, p = 0.013) and autologous hematopoietic stem cell transplant (AHCT) (22% vs. 17%, p < 0.0001) compared with AA.
Table 1.
Patient and treatment characteristics.
| Characteristics | All patients n (%) | Early mortality n (%) | No early mortality n (%) | P value | European American n (%) | African American n (%) | P value |
|---|---|---|---|---|---|---|---|
| Number of patients | 76,878 (100) | 8783 (100) | 68,095 (100) | - | 61,171 (100) | 15,707 (100) | - |
| Age at diagnosis, year, median (range) | 66 (19–90) | 73 (26–90) | 65 (19–90) | <0.0001 | 67 (19–90) | 63 (22–90) | <0.0001 |
| Age ≥ 65 years | 41,673 (54) | 6420 (73) | 35,253 (52) | <0.0001 | 34,660 (57) | 7013 (45) | <0.0001 |
| Male gender | 42,507 (55) | 5075 (58) | 37,432 (55) | <0.0001 | 34,914 (57) | 7593 (48) | <0.0001 |
| African American race | 15,707 (20) | 1612 (18) | 14,095 (21) | <0.0001 | 0 (0) | 15,707 (100) | - |
| Household annual income, zip-code based | <0.0001 | <0.0001 | |||||
| <Median ($48,000) | 33,043 (43) | 4019 (46) | 29,024 (43) | 23,025 (38) | 10,018 (64) | ||
| ≥Median ($48,000) | 42,872 (56) | 4575 (52) | 38,297 (56) | 37,358 (61) | 5514 (35) | ||
| Non-high school graduates, zip-code based | <0.0001 | <0.0001 | |||||
| <Median (13%) | 42,834 (56) | 4643 (53) | 38,191 (56) | 37,913 (62) | 4921 (31) | ||
| ≥Median (13%) | 33,119 (43) | 3955 (45) | 29,164 (43) | 22,503 (37) | 10,616 (68) | ||
| Residence | 0.38 | <0.0001 | |||||
| Metro | 61,750 (80) | 6963 (79) | 54,787 (80) | 47,929 (78) | 13,821 (88) | ||
| Urban | 11,161 (15) | 1295 (15) | 9866 (15) | 9858 (16) | 1303 (8) | ||
| Rural | 1509 (2) | 183 (2) | 1326 (2) | 1337 (2) | 172 (1) | ||
| Insurance types | <0.0001 | <0.0001 | |||||
| Private | 29,366 (38) | 1960 (22) | 27,406 (40) | 23,649 (39) | 5717 (36) | ||
| Government | 43,180 (56) | 6359 (72) | 36,821(54) | 34,567 (57) | 8613 (55) | ||
| History of prior malignancies | 11,079 (14) | 1580 (18) | 9499 (14) | <0.0001 | 9246 (15) | 1833 (12) | <0.0001 |
| Charlson comorbidity index | <0.0001 | <0.0001 | |||||
| 0 | 59,116 (77) | 5488 (62) | 53,628 (79) | 47,761 (78) | 11,355 (72) | ||
| 1 | 12,281 (16) | 1991 (23) | 10,291 (15) | 9348 (15) | 2933 (19) | ||
| ≥2 | 5481 (7) | 890 (15) | 4177 (6) | 4062 (7) | 1419 (9) | ||
| Year of multiple myeloma diagnosis | <0.0001 | <0.0001 | |||||
| 2004–2006 | 14,336 (19) | 2016 (23) | 12,320 (18) | 11,594 (19) | 2742 (17) | ||
| 2007–2009 | 18,382 (24) | 2167 (25) | 16,215 (24) | 14,675 (24) | 3707 (24) | ||
| 2010–2014 | 44,160 (57) | 4600 (52) | 39,560 (58) | 34,902 (57) | 9258 (59) | ||
| Durie-salmon stage | <0.0001 | 0.40 | |||||
| I | 1391 (2) | 25 (0.3) | 1366 (2) | 1106 (2) | 285 (2) | ||
| II | 2436 (3) | 97 (1) | 2339 (3) | 1956 (3) | 480 (3) | ||
| III | 6446 (8) | 404 (5) | 6042 (9) | 5092 (8) | 1354 (8) | ||
| Academic treatment facility | 31,939 (42) | 2687 (31) | 29,252 (43) | <0.0001 | 24,630 (40) | 7309 (47) | <0.0001 |
| Annual myeloma patient volume of facilities | <0.0001 | <0.0001 | |||||
| First quartile (0.1–3.0) | 3461 (5) | 495 (6) | 2966 (4) | 2957 (5) | 504 (3) | ||
| Second quartile (3.0–5.7) | 8657 (11) | 1296 (15) | 7361 (11) | 7352 (12) | 1305 (8) | ||
| Third quartile (5.7–10.1) | 15,726 (20) | 2326 (26) | 13,400 (20) | 12,579 (20) | 3147 (20) | ||
| Forth quartile (10.1–117.5) | 49,034 (64) | 4666 (53) | 44,368 (65) | 38,283 (63) | 10,751 (69) | ||
| Frontline treatment for multiple myeloma | |||||||
| Chemotherapy | 73,255 (95) | 8367 (95) | 64,888 (95) | 0.91 | 58,259 (95) | 14,996 (95) | 0.22 |
| Immunomodulatory imide drugs (IMiD) | 14,503 (19) | 834 (10) | 13,669 (20) | <0.0001 | 11,649 (19) | 2854 (18) | 0.013 |
| Chemotherapy and IMiD | 10,880 (14) | 418 (5) | 10,462 (15) | <0.0001 | 8737 (14) | 2143 (14) | 0.040 |
Fig. 1. Outcomes of the patients.
a Comparison of early mortality rates with 95% confidence intervals. b Multivariable logistic regression analyses for early mortality. c–e Overall survival of patients with multiple myeloma. EM early mortality.
Treatment facilities were divided into four quartiles (Q) based on annual MM patient volume. Higher AA/EA ratio was observed in larger facilities (0.27 in 3rd and 4th Q facilities) compared with smaller ones (0.18 in 1st and 2nd Q facilities) (p < 0.0001). Also, EA had longer median travel distance from their home to the treatment facility than AA; 16.1 vs. 8.1 miles for 4th Q (p < 0.0001), 8.2 vs. 5.4 miles for 3rd Q (p < 0.0001), 6.8 vs. 4.7 miles for 2nd Q (p = 0.001), and 6.9 vs. 6 miles for 1st Q (p = 0.8) facilities. On multivariable logistic regression analyses, AA had significantly lower odds of EM compared with EA (OR 0.75, 95% CI: 0.57–0.96, p = 0.027) after adjusted for all other baseline and treatment characteristics listed in Fig. 1b.
Median OS of AA vs. EA was 52.5 months (95% CI: 50.8–53.9) vs. 48.7 months (95% CI: 48.0–49.2) for all cohort (Fig. 1c, p < 0.0001), 37.8 months (95% CI: 36.6–39.0) vs. 34.9 months (95% CI: 34.3–35.5) for patients ≥65 years old (p < 0.0001), and 70.3 months (95% CI: 67.8–73.5) vs. 74.5 months (95% CI: 73.1–76.4) for patients <65 years old (p < 0.0001) (Fig. 1d).
Importantly, AA had better OS compared with EA among those aged ≥65 years at diagnosis, despite of similar utilization of IMiD (17.2 vs. 17.3%, p = 0.80) and lower utilization of AHCT (6.9 vs. 10.4%, p < 0.001) in AA. When we remove EM events for patients aged ≥65 years (i.e., the subset with favorable OS for AA) to determine the effect of EM on OS, there was no statistically significant difference in median OS between AA (45.6 months, 95% CI: 44.2–47.6) and EA (44.6 months, 95% CI: 44.0–45.4) (p = 0.08) (Fig. 1e). On multivariable Cox hazard analyses among the elderly patients aged ≥65 years at diagnosis, AA vs. EA race was not an independent predictor of OS (HR 0.89, 95% CI: 0.78–1.01, p = 0.08) after adjusted for other baseline and treatment characteristics (Supplementary Fig. 1).
Worse OS in AA compared with EA among younger patients aged <65 years at diagnosis, correlated with lower utilization of IMiD (18 vs. 21%, p < 0.0001) and AHCT (24 vs. 37%, p < 0.0001) in AA. However, removing EM events in these younger patients was not able to overcome the significantly lower median OS in AA (78.5 months, 95% CI: 75.8–81.5) compared with EA (82.2 months, 95% CI: 80.5–83.8) (p = 0.002). On multivariable Cox hazard analyses among the young patients aged <65 years at diagnosis, AA had a significantly lower risk of mortality compared with EA (HR 0.79, 95% CI: 0.68–0.93, p = 0.004) after adjusted for other baseline and treatment characteristics (Supplementary Fig. 2).
Here, we have shown that EM occurred for roughly 1 out of 7 elderly and 1 out of 15 young MM patients, therefore poses a significant challenge for continued improvement of OS in real-world practice. Among elderly MM patients aged ≥65 years at diagnosis who form 73% of EM events, AA had less EM and better OS than EA. The superior OS of AA in the elderly cohort loses its statistical significance after the elimination of EM. Also, we have noted that higher EM has contributed to lower OS in younger AA compared with younger EA, along with less novel agent use and AHCT, which highlights the young AA as a vulnerable socio-demographic subset. Overall, our findings may, in part, explain the previously reported racial disparities in outcomes of MM patients.
Racial EM disparity in elderly patients despite similar upfront induction and stage distribution can be potentially due to higher percentage of academic facilities (42 vs. 33%) and facilities with larger MM patient volume (4th Q: 64 vs. 57%) serving AA compared with EA (p < 0.0001 for all), which were independent predictors of both EM and OS along with AA race. An inherent limitation of our study is the lack of more granular data on host and disease characteristics. Assessing the role of other possible favorable factors contributing to better outcomes among elderly AA (e.g., higher bone density and lower risk of fracture in elderly AA than EA) is warranted in future studies [10].
A large population-based study previously showed that AA with MM have a younger age of diagnosis, higher incidence of and mortality from MM, greater OS and relative survival rate compared with EA [11]. This study also emphasized the difference between measures of death: OS, mortality, and relative survival rates. Higher incidence of MM in AA influences the racial mortality difference, but not OS, as mortality is dependent on incidence. In addition, relative survival rate takes the fact into account that AA have lower OS expectancy than EA in general population given the higher comorbidity burden [12].
In summary, we have described the impact of race-related EM difference on long-term OS disparity in MM patients. Our findings highlight the importance of reporting EM in explaining the discrepancy in prior reports on the outcome of MM patients. Staging systems have been developed to assess OS, but not specifically to assess the risk of EM [13–15]. A predictive model for EM has been previously proposed but not widely adapted in practice [9]. Robust strategies to target high-risk population for EM, particularly in young AA and elderly EA, may include developing risk models using disease and host factors to predict EM, promoting EM awareness in smaller cancer practices, providing equal access to novel treatment modalities and AHCT for young AA, and exploring potential race-related differences in host and tumor biology impacting EM.
Supplementary Material
Footnotes
Conflict of interest The authors declare that they have no conflict of interest.
Supplementary information The online version of this article (https://doi.org/10.1038/s41375-020-0812-2) contains supplementary material, which is available to authorized users.
References
- 1.Greenberg AJ, Vachon CM, Rajkumar SV. Disparities in the prevalence, pathogenesis and progression of monoclonal gammopathy of undetermined significance and multiple myeloma between blacks and whites. Leukemia. 2012;26:609–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Landgren O, Weiss BM. Patterns of monoclonal gammopathy of undetermined significance and multiple myeloma in various ethnic/racial groups: support for genetic factors in pathogenesis. Leukemia. 2009;23:1691–7. [DOI] [PubMed] [Google Scholar]
- 3.Al-Hamadani M, Hashmi SK, Go RS. Use of autologous hematopoietic cell transplantation as initial therapy in multiple myeloma and the impact of socio-geo-demographic factors in the era of novel agents. Am J Hematol. 2014;89:825–30. [DOI] [PubMed] [Google Scholar]
- 4.Fiala MA, Wildes TM. Racial disparities in treatment use for multiple myeloma. Cancer. 2017;123:1590–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ailawadhi S, Parikh K, Abouzaid S, Zhou Z, Tang W, Clancy Z, et al. Racial disparities in treatment patterns and outcomes among patients with multiple myeloma: a SEER-Medicare analysis. Blood Adv. 2019;3:2986–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Fillmore NR, Yellapragada SV, Ifeorah C, Mehta A, Cirstea D, White PS, et al. With equal access, African American patients have superior survival compared to white patients with multiple myeloma: a VA study. Blood. 2019;133:2615–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ailawadhi S, Jacobus S, Sexton R, Stewart AK, Dispenzieri A, Hussein MA, et al. Disease and outcome disparities in multiple myeloma: exploring the role of race/ethnicity in the Cooperative Group clinical trials. Blood Cancer J. 2018;8:67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Diversifying clinical trials. Nat Med. 2018; 24: 1779. 10.1038/s41591-018-0303-4. [DOI] [PubMed] [Google Scholar]
- 9.Terebelo H, Srinivasan S, Narang M, Abonour R, Gasparetto C, Toomey K, et al. Recognition of early mortality in multiple myeloma by a prediction matrix. Am J Hematol. 2017;92:915–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hochberg MC. Racial differences in bone strength. Trans Am Clin Climatol Assoc. 2007;118:305–15. [PMC free article] [PubMed] [Google Scholar]
- 11.Waxman AJ, Mink PJ, Devesa SS, Anderson WF, Weiss BM, Kristinsson SY, et al. Racial disparities in incidence and outcome in multiple myeloma: a population-based study. Blood. 2010;116: 5501–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cunningham TJ, Croft JB, Liu Y, Lu H, Eke PI, Giles WH. Vital signs: racial disparities in age-specific mortality among Blacks or African Americans—United States, 1999–2015. MMWR Morb Mortal Wkly Rep. 2017;66:444–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Durie BG, Salmon SE. A clinical staging system for multiple myeloma. Correlation of measured myeloma cell mass with presenting clinical features, response to treatment, and survival. Cancer. 1975;36:842–54. [DOI] [PubMed] [Google Scholar]
- 14.Greipp PR, San Miguel J, Durie BG, Crowley JJ, Barlogie B, Blade J, et al. International staging system for multiple myeloma. J Clin Oncol. 2005;23:3412–20. [DOI] [PubMed] [Google Scholar]
- 15.Palumbo A, Avet-Loiseau H, Oliva S, Lokhorst HM, Goldschmidt H, Rosinol L, et al. Revised international staging system for multiple myeloma: a report from International Myeloma Working Group. J Clin Oncol. 2015;33:2863–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.